> The first thing I will do the next time is taking more time to analyze the old system thoroughly and write a technical spec what the new system should do. A s

> The first thing I will do the next time is taking more time to analyze the old system thoroughly and write a technical spec what the new system should do. A simple "What the old system is doing now" as a spec for the rewrite does not cut it. No one knows what the old system is doing exactly so it is necessary to take the time to write a spec. I had the same experiences last year. There wasn't even a single person who knew the 'old system' on the whole. A receipt for many detours and much extra work.